GD3 ganglioside is expressed on neuroectodermal tissue and on tumors such as melanoma, sarcoma, and small-cell lung cancer. Antibodies against GD3 can shrink melanoma in rodents and patients. In clinical trials designed to immunize patients against GD3 in the adjuvant setting, we saw surprisingly low relapse rates that did not correlate with induction of anti-GD3 antibodies. Since GD3 is not presented by MHC class I or class II, we considered the alternative possibility that immunization might induce a NKT response against GD3. We have induced CD1-restricted NKT cells against GD3 in mice and in this project; we propose to characterize further the anti-GD3 murine NKT cell response. Specific Aim 1 - Define the T cell receptor usage, cytokine profile, and specificity of mouse CD1-restricted NKT cells against GD3. Most murine CD1-restricted NKT cells described to date recognize glycolipids derived from bacteria or invertebrates. Much less is known about CD1-restricted NKT cells that recognize self-glycolipids expressed on mammalian cells. Specific Aim 2 - Characterize how GD3 is loaded on to mouse CD1. In the human system, some CD1-restricted glycolipid antigens require internalization and loading in acidified late endosomes; other CD1-restricted glycolipids are loaded in post-Golgi vesicles not requiring acidification. Little is known about how self gangliosides are loaded on to CD1 although in the human system there is some evidence that GM1 ganglioside does not even require internalization to be presented. Using the mouse system, we will characterize how GD3 is loaded on to mouse CD1 assessing requirements for internalization, acidification of endosomes and Golgi function. We plan to map the intracellular processing pathway using confocal immunomicroscopy. Specific Aim 3 -Test the hypothesis that mouse NKT cells against GD3 can mediate anti-tumor effects in an antigen-specific manner. CD1-restricted NKT cells activated by alpha-galactosylceramide (from marine invertebrates) can mediate antitumor effects in a manner that does not appear to be antigen-specific. It is not known if CD1-restricted NKT cells specific for an antigen expressed on tumor cells can have anti-tumor effects. We will test the ability of CD1-restricted NKT cells against GD3 to lyse GD3+ mouse melanoma in vitro and to reject GD3+ mouse tumors in vivo. This will provide initial support for an immunotherapeutic approach based on antigen-specific NKT cells. [unreadable] [unreadable]